Table 5 in the data sheet says Approximate (NOT ensured) VCO Core Frequency Ranges. In my design there are boundaries where, as I step in frequency, the VCO core will change. I have identified them according to Table 5. This is not a problem if the VCO div doesn't change. In places where the VCO div does change and there are uncertainties on what core is chosen, how can I make sure the correct div is chosen.
As an example if I am operating at 1875M output freq with VCO4 operating at 3750M and a VCO divide by 2 (VCO_Div = 0) and I change to an output of 1880M with VCO1 and VCO divide by 1. What happens if the cal doesn't choose VCO1 for the core? What if the VCO core it chooses expects a different VCO divide than what I have selected?
Please explain what NOT ENSURED means in Table 5.
If I force the synth to use a particular core, does that mean the core range can't be guaranteed to cover the range given in Table 5? Or if I let the cal chose the core, I can't guarantee it will pick the same core as the table?.
Here are the settings I was planning to use.
| FreqMin | FreqMax | VCO Divider | VCO Core |
| 740 | <790 | 4 | VCO3 |
| 790 | <940 | 4 | VCO4 |
| 940 | <1130 | 2 | VCO1 |
| 1130 | <1355 | 2 | VCO2 |
| 1355 | <1605 | 2 | VCO3 |
| 1605 | <1880 | 2 | VCO4 |
| 1880 | <2175 | 1 | VCO1 |
| 2175 | 2500 | 1 | VCO2 |
So, I can potentially have issues at the boundaries of 940 and 1880?
How is this supposed to work?
The PLL has no way to determine what freq you are trying to program. It only multiplies the ref by (N +Nfrac/Dfrac)/VCO_divider
If the boundaries for the Cores are not stable from unit to unit then how can you make sure you have the correct VCO divider selected?
For places where the VCO change is incremental (VCO1 to VCO2 or VCO4 to VCO3) there is overlap and no change to the VCO divider and hence no issue but going from VCO4 to VCO1 there is no overlap so there could be issues with choosing VCO core and VCO divider.
